Article Text

PDF

Non-random transmission of mutant alleles to female offspring in BRCA carriers
  1. D G Evans1,
  2. A Shenton1,
  3. S Sharif1,
  4. E Woodward2,
  5. F Lalloo3,
  6. E R Maher4
  1. 1Academic Unit of Medical Genetics and Regional Genetics Service, St Mary’s Hospital, Manchester, UK
  2. 2Birmingham Women’s Hospital, Birmingham, UK
  3. 3Academic Unit of Medical Genetics and Regional Genetics Service, St Mary’s Hospital, Manchester, UK
  4. 4Division of Medical Genetics, Department of Paediatrics and Child Health, University of Birmingham, Birmingham, UK
  1. Correspondence to:
 D G Evans
 Medical Genetics and Regional Genetics Service, St Mary’s Hospital, Manchester, UK; gareth.evanscmmc.nhs.uk

Statistics from Altmetric.com

We were most interested to read the report of Gronwald and colleagues suggesting that there was non-random transmission of the mutant BRCA1 gene to female offspring in a Polish cohort.1 We have undertaken a similar analysis in an English cohort but found no evidence for non-random transmission.

Gronwald et al reported that 61.5% of unaffected female offspring carried the mutant allele and this ratio did not decrease with age (as might be expected from the exclusion of affected females at advancing age). We have analysed the ratio of positive predictive tests in first degree relatives of proven mutation carriers from 284 BRCA1 and BRCA2 families from central and north western England. Like Gronwald et al1, we excluded women who were affected with cancer or whose daughter was affected as they were potential obligate carriers. We did not include branches of the family where the mutation status of the parent was unknown even if it was a mother with breast or ovarian cancer in a family with a known mutation. This excluded the possibility of including the testing of daughters of potential phenocopies. We did not reach the level of 89% uptake of predictive testing achieved with the daughters of 91 female mutation carriers in the Polish study. This was presumably due to the different approaches to offering testing. No results were available for females <18 years old in our study. Mutation testing was entirely at the behest of the unaffected female.

The overall frequency of positive tests was 45% in our series (table 1) and, in contrast to the Polish study, there was a sharp decline in the ratio of positive to negative tests in unaffected women after 40 years of age. Whereas 61.5% of women tested aged <50 years in Poland were positive, the proportion was only 51% in our series. However, intriguingly, the proportion of positive results was substantially higher than 50% in the 30–39 year old cohort suggesting some replication of the results of the Polish study, although these results did not reach statistical significance (p = 0.29). It is possible that women with breast symptoms may be more likely to come forward for testing and then subsequently be diagnosed with breast cancer. Three carriers and three mutation negative women in our series have since developed breast cancer, none of whom had breast symptoms at the time of testing. There was no difference in the results for BRCA1 compared to BRCA2 with 45% testing positive for each gene and no evidence of a difference in age related change. Taking the results from our survey together with those from two similar studies,2,3 the results from women <50 years of age reveal that 54% were mutation positive. There was a significant excess of mutation carriers <30 years of age in two previous studies.2 However, this excess disappears (table 2) when the results of these two studies are added to those of our study and another study (57% positive; p = 0.2).3

Table 1

 Proportion of unaffected female first degree relatives of BRCA1/2 mutation carriers testing positive for the family mutation

Table 2

 Proportion of unaffected female first degree relatives of BRCA1/2 mutation carriers testing positive for the family mutation from all four studies

We suspect that differences in the ascertainment strategies may explain the differences between the findings in Poland and the UK. Thus, in the UK families penetrance appears to be high with <33% of unaffected females older than 50 years of age testing positive (as would be expected with a 50% penetrance by that age).1,4,5 Indeed, adding the results of the other two studies, only 30% of women >50 years of age were mutation positive. Although there was a very low level of positive tests in those over 60 years of age, we had excluded seven living affected female mutation carriers whose daughters had developed breast or ovarian cancer. The UK families were nearly all selected on the basis of having at least three (and usually four) or more affected relatives with breast/ovarian cancer. In contrast, the Polish families were much less highly selected because the presence of common founder mutations in the Polish population allows a lower threshold for initiating BRCA mutation analysis. Although we have offered testing more widely for two common mutations in BRCA1/2 in Manchester, the penetrance in these families was still high.6 Family and epidemiological studies have demonstrated that approximately 70–80% of BRCA1 and BRCA2 mutation carriers develop breast cancer in their lifetime, although the risk is a little lower for BRCA2.4,7,8 The very low figures published on small numbers of families from population studies have now been addressed by a meta analysis.8 An important practical point is that penetrance estimates provided in genetic counselling sessions should be modified according to how the family was ascertained. Thus, families which are similar to those from the linkage consortium2 should be provided with higher risk estimates, whereas those ascertained with fewer affected relatives should be quoted a range including the lower estimates derived from population based studies.9–11

Uptake of genetic testing is age related, with high uptake in unaffected females in their thirties and forties and lower uptakes outside this age group. In particular the lower uptake in older women reflects the reduced interest in testing by those at risk women without daughters who may feel they are past the personal risk period for their family.12 We did include results from daughters who had a father who carried the mutation although this represents <10% (29) of the tested females. The results were not altered by excluding this subset (48% tested positive); although de la Hoya et al2 did show a lower proportion of carriers born to carrier fathers, this was with only 59 tested daughters. In summary, although our findings differ from those of Gronwald et al,1 we suggest that the hypothesis that there may be a selective advantage for the BRCA1 mutant allele in utero merits further investigation and that further studies should also include BRCA2 families and possibly also those who inherited the mutation from their father.

References

View Abstract

Footnotes

  • We thank Cancer Research UK (Cancer Genetics Network) for financial support.

  • Conflicting interests: none declared.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.